What If the TV Was Off? Examining Counterfactual Reasoning Abilities of Multi-modal Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Letian, Zhai, Xiaotong, Zhao, Zhongkai, Zong, Yongshuo, Wen, Xin, Zhao, Bingchen
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929314760491008
author Zhang, Letian
Zhai, Xiaotong
Zhao, Zhongkai
Zong, Yongshuo
Wen, Xin
Zhao, Bingchen
author_facet Zhang, Letian
Zhai, Xiaotong
Zhao, Zhongkai
Zong, Yongshuo
Wen, Xin
Zhao, Bingchen
contents Counterfactual reasoning, a fundamental aspect of human cognition, involves contemplating alternatives to established facts or past events, significantly enhancing our abilities in planning and decision-making. In light of the advancements in current multi-modal large language models, we explore their effectiveness in counterfactual reasoning. To facilitate this investigation, we introduce a novel dataset, C-VQA, specifically designed to test the counterfactual reasoning capabilities of modern multi-modal large language models. This dataset is constructed by infusing original questions with counterfactual presuppositions, spanning various types such as numerical and boolean queries. It encompasses a mix of real and synthetic data, representing a wide range of difficulty levels. Our thorough evaluations of contemporary vision-language models using this dataset have revealed substantial performance drops, with some models showing up to a 40% decrease, highlighting a significant gap between current models and human-like vision reasoning capabilities. We hope our dataset will serve as a vital benchmark for evaluating the counterfactual reasoning capabilities of models. Code and dataset are publicly available at https://bzhao.me/C-VQA/.
format Preprint
id arxiv_https___arxiv_org_abs_2310_06627
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle What If the TV Was Off? Examining Counterfactual Reasoning Abilities of Multi-modal Language Models
Zhang, Letian
Zhai, Xiaotong
Zhao, Zhongkai
Zong, Yongshuo
Wen, Xin
Zhao, Bingchen
Computation and Language
Computer Vision and Pattern Recognition
Machine Learning
Counterfactual reasoning, a fundamental aspect of human cognition, involves contemplating alternatives to established facts or past events, significantly enhancing our abilities in planning and decision-making. In light of the advancements in current multi-modal large language models, we explore their effectiveness in counterfactual reasoning. To facilitate this investigation, we introduce a novel dataset, C-VQA, specifically designed to test the counterfactual reasoning capabilities of modern multi-modal large language models. This dataset is constructed by infusing original questions with counterfactual presuppositions, spanning various types such as numerical and boolean queries. It encompasses a mix of real and synthetic data, representing a wide range of difficulty levels. Our thorough evaluations of contemporary vision-language models using this dataset have revealed substantial performance drops, with some models showing up to a 40% decrease, highlighting a significant gap between current models and human-like vision reasoning capabilities. We hope our dataset will serve as a vital benchmark for evaluating the counterfactual reasoning capabilities of models. Code and dataset are publicly available at https://bzhao.me/C-VQA/.
title What If the TV Was Off? Examining Counterfactual Reasoning Abilities of Multi-modal Language Models
topic Computation and Language
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2310.06627